A Robustness-aware LEO Satellite Networking and Resource Allocation Strategy

Different from the traditional edge computing network architecture, LEO satellite constellations are resource-constrained and highly dynamic, making it difficult for traditional edge computing network models to be applied to LEO satellite constellation-based task offloading environments directly. In addition, current works on interstellar networking for LEO satellites rarely consider the impact of missing interstellar links on networking robustness due to frequently changing interstellar topology, which obviously degrades the performance of interstellar collaboration. Based on this, we study the cooperative networking and resource allocation mechanism of LEO satellites with the goal of maximizing the networking computational capability and networking robustness. We propose an approach based on greedy strategy and McCormick envelope theory, namely GREEDY-LR to solve the above problem efficiently. Finally, simulation experiments prove the effectiveness and advantages of the algorithm proposed in this paper.